Identification and Characterization of Gliocladium viride Isolated from Mushroom Fly Infested Oak Log Beds Used for Shiitake Cultivation.

A green mold species that has not previously been reported in Korea was isolated from oak log beds used for shiitake (Lentinula edodes) cultivation that were infested by mushroom flies. In this study, we identify the mold species as Gliocladium viride (an anamorph of Hypocrea lutea) and describe its mycological properties. The fungus was cottony on both potato dextrose agar (PDA) and Czapek yeast extract agar (CYA), but was colored white on PDA and became yellowish green and brown on CYA. Mycelial growth on PDA attained a diameter of 73 mm at 30℃ after 5 days. The fungus grew faster on malt extract agar (> 80 mm, 5 days at 25℃) compared to CYA and PDA (< 68 mm, 5 days at 25℃). Penicillate conidiophores of the fungus are hyaline, smooth walled, branching above typically in four stages, and 120~240 µm in length. Club-shaped or slender phialides are formed on the metulae. Conidia of the fungus were ovate and elliptic, yellowish brown and green, and 2.5~3.0 µm × 1.8~2.3 µm in size. Typically, slimy conidia are formed in a mass and colored brown to dark green to almost black. The internal transcribed spacer rDNA and translation elongation factor 1 alpha gene sequences of the fungus isolated here show 99% identity with previously identified G. viride strains.

Shiitake is a popular edible and medicinal mushroom in Korea. To produce high quality mushrooms, shiitake has been cultivated on oak log beds. Recently, many farmers in the shiitake mycoculture industry in Korea have been stressed by severe yield losses due to damage of mushroom log beds caused by infestations of mushroom flies. So far, the identity of the infesting mushroom flies has not yet been elucidated, despite them being frequently found in damaged oak log beds of shiitake on many mushroom farms across Korea. It has been postulated that the damage to the mushroom log beds are caused not only by mushroom files, but also by green fungi. Some species belonging to the Hypocrea, Trichoderma, and Gliocladium genera have been known as agents of green mold disease, which affects cultivated mushrooms such as Agaricus bisporus, Lentinula edodes and Pleusotus ostreatus [1, 2]. However, no investigation has yet determined what kinds of green molds are present and which species are problematic in damaging shiitake log beds infested by mushroom flies. Therefore, we have examined green molds in the mushroom fly-infested shiitake log beds. In this study, we report the isolation and identification of Gliocladium viride, which has not been described before in Korea.

Materials and Methods

Fungal isolation

Samples of damaged oak log beds were obtained in the summer of 2009 from mushroom farms in Cheonan city, Korea (Fig. 1). The sampled log beds were split into small pieces (0.5 cm × 0.5 cm × 0.2 cm), surface sterilized in 1% sodium hypochlorite solution for 2 min, washed three times with sterile water and put on potato dextrose agar (PDA) containing streptomycin (100 µg/mL) and incubated for several days at 25℃. Mycelia grown out from the small pieces were transferred to fresh PDA and single spore isolates were obtained from the PDA-grown fungi. The obtained single spore isolates were maintained on PDA for the duration of the experiment and stored either at -80℃ in 10% glycerol for long-term storage or at 4℃ for short-term storage.

Fig. 1

Examples of damaged shiitake log beds infested by mushroom flies. Arrow (A) and (B) indicates the spot where green molds were isolated in this study.

Morphological studies

Colony appearance and sporulation were tested from cultures grown in darkness at 25℃ for 5~7 days on PDA, Czapek yeast extract agar (CYA) and malt extract agar (MEA). The growth rates of the isolates were determined at 15℃, 20℃, 25℃, 30℃ and 37℃ on PDA for 5 days. Agar disks (5 mm diameter) taken from the edge of an actively grown colony were placed at the end of a PDA plate. Three replicate plates were tested for each temperature. Light microscope observations and mycelia length measurements were made in 0.05% KOH solution by a phase contrast Axioskop 40 (Karl Ziess, Oberkochen, Germany). Mycelial measurements are reported as the maximum and minimum values of 30 measurements, given between brackets, as well as the mean ± standard deviation. For scanning electron microscopic observations, the fungal isolates were grown for 3~5 days at 25℃ on PDA. Agar blocks were fixed with 2% glutaraldehyde in 0.1M cacodylate buffer for 16 hr and 1% osmium tetroxide in 0.1M phosphate for 1 hr. The fixed samples were subsequently washed with 0.05 M cacodylate buffer, dehydrated in a series of ethanol washes (50% for 20 min, 75% for 20 min, 90% for 20 min, 95% for 20 min and 100% for 20 min), passed through ethanol-isoamyl acetate, dried with a Hitachi critical point dryer and coated with platinum-palladium at 25 nm using an Hitachi E-1030 ion sputter (Hitachi Science Systems, Ltd., Tokyo, Japan). The prepared specimens were examined by a Hitachi S-4200 scanning electron microscope (SEM) operating at 10 kV.

DNA extraction, PCR amplification and DNA sequencing

Fungi were grown for 3~5 days on PDA at 25℃ and fungal genomic DNA for PCR was obtained from the mycelia of the PDA-grown cultures using the drilling method described by Kim et al. [3]. The internal transcribed spacer (ITS) ribosomal DNA regions were amplified by PCR using the universal primer pair, ITS1 and ITS4 [4]. The tef1-α gene, encoding translation elongation factor 1-alpha, was amplified using the primers TEF728 and TEF1 [5]. Sequencing was performed on an ABI 3700 automated sequencer (Perkin-Elmer Inc., Waltham, MA, USA) at the DNA synthesis and sequencing facility, MACROGEN (Seoul, Korea). The determined nucleotide sequences were then compared to publicly available sequences by BLASTN against the GenBank database (http://www.ncbi.nlm.nih.gov/BLAST) and by Tricho-BLAST at the website of the International Subcommission on Trichoderma and Hypocrea Taxonomy (ISTH, http://www.isth.info/). ISTH is a publicly available database with sequence diagnosis and similarity search tools, which covers all genetically characterized species of the Trichoderma and Hypocrea genera.

Molecular phylogenetic analysis

The determined nucleotide sequences were manually edited using the Chromas v2.31 program (Technelysium Pty. Ltd., Helensvale, Qld, Australia) and aligned using the ClustalW2 program (European Bioinformatics Institute, Cambridge, UK). Reference sequences of related taxa were obtained from the GenBank database. The aligned sequences were analyzed with the PAUP 4.0 b10 program [6]. Phylograms based on ITS and tef1-α gene sequences were constructed by the neighbor-joining method [7]. Bootstrap values were generated with 1,000 replicates through heuristic searches. Fusarium solani was used as an outgroup.

Results and Discussion

Colony morphology

Initially several fungi were isolated from the damaged oak log beds (Fig. 1). When they were grown on PDA and MEA, they showed similar morphological properties. Thus, three of them were selected, coded as DKU002, DKU003 and DKU004, and used for this study (Table 1). Among these three isolates, DKU002 was used as a representative isolate for morphological description. The colony morphology and color of DKU002 grown on three different media are given in Fig. 2. The fungus was cottony on both PDA and CYA, but was colored white on PDA and became yellowish green and brown on CYA. The bottoms of the fungal cultures, seen on the reverse side of the plates, were colorless or light brown. Colony color was clearer on CYA than on MEA and PDA. On all three media, the fungus produced thin, transparent hypae and its mycelia broadly extended over the entire medium. Mycelia were more densely formed on PDA and CYA than on MEA. These observations show that colony colors and mycelia growth patterns of DKU002 vary depending on type of media. On MEA, mycelia were flat with few aerial mycelia. Weak, unevenly spaced concentric rings resulted from different growth rate of mycelia also formed on MEA, but were not observed on PDA or CYA plates. On MEA a greenish color appeared along with concentric rings. On PDA and CYA, the green color mostly appeared at the edge of the plates. DKU002 was a rapidly growing fungus, and grew faster on MEA and PDA than on CYA (Fig. 3). After 5 days of incubation on PDA, mycelia of the fungus grew to a diameter of 16 mm at 15℃, 39 mm at 20℃, 68 mm at 25℃, 73 mm at 30℃ and 25 mm at 37℃ (Fig. 3). These results show that the fungus grows well at 25~30℃ at which shiitake cultivation normally occurs.

Table 1

Morphological characters of Gliocladium viride and the isolate DKU002

aDescription referred from G. deliquescens, the synonym of G. viride [8].

Fig. 2

Colony patterns of DKU002 grown on potato dextrose agar (A), Czapek yeast extract agar (B) and malt extract agar (C) after 7 days of culturing at 25℃.

Fig. 3

Variations in mycelial growth of DKU002 on PDA at different temperature (A) and on different media at 25℃ (B). CYA, Czapek yeast extract agar; MEA, malt extract agar; PDA, potato dextrose agar.

Morphological characters

Light and scanning electron microscopic images of DKU002 are given in Fig. 4. The fungus has an erect penicillate conidiophore structure (Fig. 4O and P). Conidia formed without chains at the apices of the conidiophores (Fig. 4E and K). These microstructures show characteristics of the genus Gliocladium (Fig. 4G~I and K). Conidiophores of the fungus are hyaline, smooth walled, typically upper branching, and 120~240 µm in length (Table 2). Conidiophores arise from both submerged and surface hyphae, with several growing from each point; both aerial and submerged stolons are present at these points (Fig. 4A~D). Conidia fructification is shown typically in four stages (Fig. 4N~P). Three to five primary branches (ramus) arise from the conidiophore stipe. Each primary branch bears a verticil of secondary branches, namely, verticils of metulae. Club-shaped or slender phialides are formed on the metulae (Table 2). The shape of the primary and secondary branches and metulae are elongate and oblong. Overall, each series of branches in the conidiophores are progressively smaller. Conidia were ovate and elliptic, yellowish brown and green, and 2.5~3.0 µm × 1.8~2.3 µm in size (Table 2, Fig. 4F and Q). Conidia production is very abundant on CYA, usually enveloping the entire colony. Typically, slimy conidia are formed in a mass ranging in color from brown to dark green to almost black (Fig. 4B~E and K). Chlamydospore-like structures are seen (Fig. 4N). The conidiophore base has extending root-like (rhizoid-like) hyphae (Fig. 4 and M). Perithecia and sclerotia were not found.

Fig. 4

Morphological features of DKU002 observed by a light microscope (A~N) and scanning electron microscope (SEM) (O~Q). Conidiophores and wet slimy conidia masses formed on potato dextrose agar (A~C) and on malt extract agar (D~E). Conidia (F, Q), conidiophores (G~K), an extending root-like structure (L~M) and a chlamydospore-like structure (N) were all observed. Penicillate with four stage branches (O~P) were observed under SEM.

Table 2

Glicladium viride isolates used in this study

ITS, internal transcribed spacer.

According to the taxonomy key of the Gliocladium genus, species having conidial areas that are pale yellow-green to dark green belong to the G. catenulatum series or the G. deliquescens series [8]. If conidia are typically pale yellow-green and commonly remain in chains to form wet columns, the species belongs to G. catenulatum series. However, if the conidia are typically dark green to almost black and collect into slime balls, the species belongs to G. deliquescens series. In this study, the observed morphological characteristics of DKU002 were similar to those of the G. deliquescens series, which includes G. deliquescens, G. atrum and G. nigro-virescens. Among the species in the G. deliquescens series, the description of G. deliquescens most matched the morphological characteristics of DKU002. The comparison of G. deliquescens and DKU002 is given in Table 2. G. deliquescens is a synonym of Gliocladium viride [9], which is the anamorph of Hypocrea lutea [10].

For further evidence to confirm whether DKU002 is the anamorph of H. lutea, its ITS rDNA and tef1-α gene sequences were analyzed. We amplified these target DNA sequences by PCR and determined the nucleotide sequences of the PCR amplicons. 573 (ITS rDNA) and 637 (tef1-α) bp nucleotide sequences were determined from DKU002, DKU003, and DKU004, respectively. These three isolates shared 99% sequence identity in both the ITS rDNA and tef1-α genes. These results indicate that the three DKU isolates are the same species. When homologous sequences for these ITS rDNA sequences were mined from the GenBank DNA database, the experimentally determined sequences most matched (99% similarity) the ITS rDNA sequence of Gliocladium deliquescens (accession number GQ229478), Gliocladium viride (accession number EU076919), and Hypocrea lutea (accession number AB027384). In the case of the tef1-α gene, the sample sequences most matched (99% similarity) the tef1-α gene sequences of H. lutea (accession number FJ860772 and FJ860644). Phylogenetic analyses based on the ITS region (Fig. 5A) and tef1-α gene (Fig. 5B) sequences also placed DKU002~004 with H. lutea (telemorph of G. viride). When we searched the ITS rDNA and tef1-α gene sequences of DKU002 through TrichoBLAST, the sequences matched with those of H. lutea and H. melanomagnum. Chaverri and Samuels [10] noted that the anamorph of H. melanomagnum (Trichoderma melanomagnum) is almost indistinguishable from the anamorph of H. lutea (Gliocladium viride). In their description of H. melanomagnum/Trichoderma melanomagnum, no chlamydospore was observed and colonies grown on PDA showed concentric rings with regular intervals. But in this study, DKU0002 produced chlamydospore-like structures and did not produce concentric rings on PDA. Thus, we could differentiate DKU002 from H. melanomagnum/Trichoderma melanomagnum. Our morphological and molecular data conclusively suggest that the green mold species from shiitake log beds is G. viride

Fig. 5

Phylogenetic relationships of the DKU isolates to other related species. Cladograms based on analysis of a partial nucleotide sequence of the internal transcribed spacer region (A) and the tef1-α gene (B) was generated by neighbor-joining analysis. Numbers at nodes represent the percentage of 1,000 bootstrap resampling runs that result in the same branching. Fusarium solani was used as an outgroup.

Overall, we identified the anamorph of H. lutea and described its morphological characteristics and its molecular and physiological properties. This is the first description of Gliocladium viride in Korea. This fungus is usually abundant in soils. Recently, it has been isolated from mushroom cultivation facilities in Japan (GenBank accession number AB298706). However, the significance of this fungus in mushroom cultivation has not yet been identified. Thus, further properties of this species need to be explored, especially in Korea, to determine whether it plays a role in the mushroom fly associated deterioration of shiitake oak log beds.